Assessing the status of the flooding and flood risk by using Analytical Network Process-fuzzy logic hybrid model(Case study: lighvan River Basin)

Flood is largest and most important climate crisis that kills thousands of people every year. And causes many damages to human society and the environment. This phenomenon of the distant past has always been a fear in human society (Nasrin Nejad et all, 2014: 16). Flood is one of the three natural disasters and major events in Iran and to note that at least a year, a major flood occurred in the country. According to studies in each year number of 40 small and large floods occurs in parts of the country (Department of Planning And Strategic Supervision president, 2010). Today is considered Non-structural deal with floods in catchment. Accordingly, the first action that comes to reduce the threat of flooding is flood control in the upper basin (Abdi, 2006:200). One of the common methods in combating the floods is Identifying in terms of potential flooding in basin. zoning possibility of recognizing watershed Specifies in terms of the flood.
Several studies have been conducted in the field of flood inside and outside the country that among these can be noted in research conducted by Thilagavathi et all(2011), Ozturk et all(2013), nasrin Nejad et all (2014), eskandary nejad et all (2015) and etc. Determination of flooding and flood-prone zoning Lighvan River is the main objective of this research. Lighvan River originates from the northern slopes of Sahand and after passing through the Tabriz and Basmenj cities eventually pours into Urmia Lake.
Determination of flooding and flood-prone zoning Lighvan River is the main objective of this research. To do so has been used a combination of two Analytical Network Process and fuzzy logic models and along with 12 environmental parameters.
Analytical Network Process: Analytical Network Process model is broad form of AHP model, Analytical Hierarchy Process is introduced by the Saaty (1980). The underlying assumption of this method is Existence of independent sub-criteria together (Saaty, 2006). Saaty In cases where this principle is violated, and structure of issue has become a network, introduces Analytical Network Process model (Amalnik et al, 2010:202). Analytical Network Process for each subject considers in the form of a network of criteria, sub-criteria and options that have been gathered together in clusters. This model has several steps, which are as follows: 1) build models and create a network structure 2) binary comparison and determine the priority vectors - create a unweight super matrix 4) Creating weighted super matrix 5) finally, create a Limit super matrix.
Fuzzy logic model: Fuzzy logic model is a generalization of the classical set theory in mathematics, science, and every day is a new method to express uncertainty and ambiguity. Fuzzy sets are defined by membership functions. For each fuzzy set between zero and there is a lack of full membership and a full membership represents zero (Hosseini et al, 1390). Fuzzy modeling is done by using several operators. An important function of the fuzzy logic model can be fuzzy Algebraic multiplication operator (Fuzzy Product), sum Fuzzy (Fuzzy Sum) gamma phase and so on. This operator of multiplication by the algebraic sum of fuzzy based on the fuzzy relation (1) is defined. Equation (1)
The parameters used in the modeling, due to the nature of their were categorized in four clusters such as: hydroclimate, geomorphology, geology and land cover for potential zoning for the flooding and two clusters of geomorphology and hydroclimate for zoning. Continue on with taking the network connection between the clusters and the parameters, established network structure between the parameters. By forming network structure between the clusters and the parameters arrives stage implementation of Analytical Network Process model, the formation of super matrix's and paired comparison. Due to the existence of two goals in the research (flood and flooding potential zoning) and impact of different parameters in them, Analytical Network Process model was conducted for each of the goals individually and in view of Effect In the flooding and flood risk and their relative weights Was obtained to each other.
Based on Comparisons of currently criteria, slope (0.18), lithology (0.125), aspect (0.118) have highest impact factor in flooding and slope parameters (0.229), the plan curvature (0.2 ) and profile curvature (0.134) have the highest impact factor in the flood risk in the study basin.
According to this research results, the southern part of the basin have been introduced as the prone to flooding areas with classifying in zones with high and very high potential. This area due to volcanic bedrock, Low permeability, High slope, more rainfall and high drainage density Have the ability to produce high runoff And that is why are located in these classes. In terms of flood risk the northern half of the basin's the greatest potential for flood risk. These areas mainly include land surrounding the main River, low and flat lands portion of the basin output. In terms of subdivisions land units these regions, mainly are known as the land with low roughness and relative slope less than 10% that included marginal areas of rivers, old and new terraces And have been characterized With features such as low slope, low relative height, low distance from river and the concave plan and profiles curvature.